In an existing point-to-multipoint (p2mp) ring network protection method, usually a closed-loop channel in a direction opposite to each working channel is established as a protection channel for each working channel. The allocation of labels for the protection channel must be associated with the allocation of labels for the working channel, so as to ensure LSP (label switching path)-based protection switching for a p2mp service between the working channel and the protection channel.
FIG. 1 is a schematic view showing an existing p2mp ring network protection method. As shown in FIG. 1, a solid line without any arrow represents a physical link, a solid line with an arrow represents LSP for a p2mp service, and a dotted line with an arrow represents an established protection channel. As can be seen from FIG. 1, the p2mp service enters the ring at a node A, while it leaves the ring at nodes D and E. Accordingly, it is required to establish two protection channels for the p2mp service.
FIG. 2 is a schematic view showing a situation where there are two p2mp services in one ring. As shown in FIG. 2, in addition to the p2mp service shown in FIG. 1, another p2mp service enters the ring at a node F, while it leaves the ring at nodes D and E. Although this p2mp service and the p2mp service shown in FIG. 1 have same off-ring nodes, they have different on-ring nodes and thus another two protection channels are required to be established for this p2mp service.
It can therefore be seen that, in the related art, the protection channels are established by taking each p2mp service which enters the ring as a unit. In other words, the number of the protection channels established for each ring depends linearly on the number of the p2mp services. As a result, there are too many protection channels, and the resultant network structure is very complex.